Changes in Body Composition with Weight Loss - Wiley Online Library

13 downloads 5176 Views 174KB Size Report
May 5, 2007 - fat-free mass, total bone mineral content, total body protein, and body cell mass at ... Discussion: Weight loss programs used in this study in- duced fat loss without ... E-mail: [email protected]. Copyright © 2007 ...
Changes in Body Composition with Weight Loss: Obese Subjects Randomized to Surgical and Medical Programs John B. Dixon,* Boyd J. G. Strauss,† Cheryl Laurie,* and Paul E. O’Brien*

Abstract DIXON, JOHN B., BOYD J. G. STRAUSS, CHERYL LAURIE, AND PAUL E. O’BRIEN. Changes in body composition with weight loss: obese subjects randomized to surgical and medical programs. Obesity. 2007;15: 1187–1198. Objective: To assess changes in body composition with weight loss in obese subjects randomized to a laparoscopic adjustable gastric band surgical program or a medical program using a very-low-energy diet and orlistat. Research Methods and Procedures: Using body composition measurements by DXA, neutron activation for total body nitrogen, and whole body ␥ counting for total body potassium, we studied changes in fat mass, fat distribution, fat-free mass, total bone mineral content, total body protein, and body cell mass at 6 (n ⫽ 61 paired) and 24 months (n ⫽ 53 paired) after randomization. Results: At 24 months, the surgical group had lost significantly more weight (surgical, 20.3 ⫾ 6.5 kg; medical, 5.9 ⫾ 8.0 kg). There was favorable fat-free mass to fat mass loss ratios for both groups (surgical, 1:5.5; medical, 1:5.9). Changes in total body nitrogen or potassium were favorable in each group. A small reduction in mean bone mineral content occurred throughout the study but was not associated with extent of weight loss or treatment group. At 6 months, weight loss for both groups was similar (surgical, 14.1 ⫾ 4.5 kg; medical, 13.3 ⫾ 7.3 kg). The medical program subjects lost less fat-free mass and skeletal muscle

Received for review March 16, 2006. Accepted in final form November 24, 2006. The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. *Centre for Obesity Research and Education (CORE), Monash University, Melbourne, Victoria, Australia; and †Body Composition Laboratory, Monash Medical Centre, Clayton, and Monash University Department of Medicine, Victoria, Australia. Address correspondence to John Dixon, Centre for Obesity Research and Education, Monash Medical School, The Alfred Hospital, Melbourne, Victoria, Australia 3004. E-mail: [email protected] Copyright © 2007 NAASO

and had increased total body protein. The proportion of body fat to limb fat remained remarkably constant throughout the study. Discussion: Weight loss programs used in this study induced fat loss without significant deleterious effects on the components of fat-free mass. Key words: gastric band, fat-free mass, DXA, nitrogen, potassium

Introduction Weight loss is the most logical, and is proving to be the most effective, way of treating the host of diseases associated with overweight and obesity. Many of the metabolic and inflammatory processes associated with central obesity are attenuated with weight loss, and diseases such as type 2 diabetes and obstructive sleep apnea are dramatically improved or resolved (1). In addition, there are major improvements in quality of life, body image, and psychologic status (1). The method of weight loss may also be important. To date, dietary and drug therapies have been of limited value in achieving and sustaining significant weight loss (2,3). On the other hand, bariatric surgical therapy has proven to be very effective at achieving and sustaining weight loss in at least the medium term (3 to 5 years) (4,5). Although the benefits of weight loss have been clearly shown, the risks of weight loss have not always been so actively studied. Discussion of side effects and complications associated with therapy for obesity has usually focused on the therapy itself, rather than the outcomes of weight loss on body composition. Examination of body composition changes with weight loss is useful in comparing the specific therapies involved and in addressing general concerns that may follow significant weight loss in humans (6,7). In a recently published randomized controlled trial of medical therapy for weight loss, in which an intensive program of behavioral therapy, very-low-energy diet OBESITY Vol. 15 No. 5 May 2007

1187

Body Composition with Weight Loss, Dixon et al.

(VLED),1 and orlistat was compared with bariatric surgery using the laparoscopic adjustable gastric band (LAGB), we showed a similar weight reduction in each group at 6 months after randomization, but by 2 years, the surgical group had achieved and sustained greater weight loss and better health (8). All subjects had a BMI between 30 and 35 kg/m2 on entry into the study. Both groups achieved significant weight loss during the study period, with a loss of 85% of excess weight in the surgical group and 21% of excess weight in the medical group. The surgical group also showed a significant reduction in the prevalence of the metabolic syndrome and improved health-related quality of life. This randomized controlled trial of two methods of weight loss provided us with an opportunity to examine in detail changes in body composition during the 2-year study period. Using basic anthropometric measures, total body DXA, total body nitrogen, and total body potassium analyses, we assessed changes in fat mass, fat distribution, fatfree mass, bone density, total bone mineral content, total body protein, and body cell mass at 6 and 24 months after randomization. In addition, changes in reported physical activity were assessed for associations with changes in weight, fat-free mass, bone mineral content, and distribution of regional fat stores.

Research Methods and Procedures Participants Patients were recruited for this randomized trial through an advertisement in the newspaper. All patient assessments and outpatient treatments were conducted at a community clinic dedicated to obesity management or in the clinics of a university department of surgery. Surgical procedures were conducted in a hospital experienced in the care of bariatric surgical patients. The study was approved by the Human Ethics committees of The Alfred Hospital and The Avenue Hospital in accordance with the guidelines of the National Health and Medical Research Council and with the Helsinki Declaration of 1975 (as revised in 2000). Patients were considered eligible if they were between 20 and 50 years of age, had a BMI of 30 to 35 kg/m2 at the initial examination with identifiable problems, either medical, physical, or psychosocial, associated with their obesity, gave a history of attempts at weight reduction over at least a 5-year period, were able to understand the options offered and the randomization process, and were willing to comply with the requirements of each program. Details of the exclusion criteria, sample size calculations, and randomization process have been published in an earlier report.

1 Nonstandard abbreviations: VLED, very-low-energy diet; LAGB, laparoscopic adjustable gastric band; CI, confidence interval.

1188

OBESITY Vol. 15 No. 5 May 2007

Interventions Surgical: LAGB. The LAGB (LAP-BAND System; Inamed Health, Santa Barbara, CA) procedure was performed by a standardized method (9) by two surgeons experienced in the technique, within 1 month of randomization. Patient progress was reviewed by the treating surgeon every 4 to 6 weeks throughout the study period, and adjustments to the volume of saline within the band were made using standard clinical criteria (10). The surgical program provided regular advice regarding eating patterns and behaviors, dietary quality and balance, multivitamin supplementation, and physical activity. This program was provided by the same multidisciplinary team, in the same manner, and blended with the program provided for a much larger cohort of patients being followed after LAGB surgery. The plan was to provide a usual care bariatric surgical program. Medical: VLED/Orlistat. This medical program centered on the use of behavioral modification, VLED, and pharmacotherapy, with education and professional support regarding appropriate eating and exercise behavior being provided periodically by trained physicians, dietitians, and research assistants. The aim was to achieve significant weight loss over an initial intensive 6-month period. A VLED (Optifast; Novartis Consumer Health Australasia, Mulgrave, Australia) was undertaken for 12 weeks, followed by a transition phase over 4 weeks that combined VLED with normal meals plus Orlistat 120 mg (Xenical; Roche Laboratories, Nutley, NJ) before each normal meal, after which Orlistat 120 mg was taken before all meals until the completion of the intensive 6-month phase. This initial 6-month program was followed by continual behavioral, dietary, and exercise advice designed to assist the participant in maintaining weight loss over a prolonged period. Some individualization of the management program was based on the judgment of the treating physician and medical consultants so that management reflected good clinical practice. The physician saw each patient every 2 weeks during the intensive program and on an every 4- to 6-week basis throughout the rest of the study. Body Composition Measures All measurements were performed at the Body Composition Laboratory, Clinical Nutrition and Metabolism Unit, Monash Medical Centre, Melbourne, Australia. Serial assessments were performed using the same equipment. Two experienced nurses using standard methods took anthropometric measures of weight (to the nearest 0.1 kg) on a regularly calibrated scale, height (to the nearest 0.1 cm) using a Harpenden stadiometer (Holtain Ltd., Crymych, UK), and abdominal and gluteal circumferences at the sites defined by the World Health Organization. Data calculated from these measurements included BMI and waist-to-hip ratio.

Body Composition with Weight Loss, Dixon et al.

Total body DXA was performed using a Lunar DPX absorptiometer (GE Healthcare, Madison, WI) in slow mode, and data were analyzed using software version 3.6z (Lunar Corp., Madison, WI). Data collected included total body bone mineral content and density, which were compared with Australian reference data, and total body fat mass, fat-free mass, and regional fat, fat-free mass, and skeletal muscle mass (11). Total body potassium was measured using a whole body shadow shield ␥ counter, with four NaI crystal detectors, arranged two above and two below the subject, with counting taking place over 1000 seconds. The method uses a phantom of known potassium content as a reference. Normalized total body potassium has been developed using a predictive equation (12). Body cell mass is calculated from total body potassium using the following formula: body cell mass ⫽ 0.0092 ⫻ total body potassium (13). Total body protein was measured using prompt ␥ in vivo neutron activation analysis for nitrogen, using a cf (252) neutron source and four NaI detectors arranged in pairs on either side of the patient with counting taking place over 1000 seconds. Total body protein was calculated as 6.25 ⫻ body nitrogen content. Calculated data derived from nitrogen measurement also included a nitrogen index derived from comparison of the measured total body nitrogen to the expected total body nitrogen from a healthy reference population matched for sex, age, and height (14). All patients were asked about physical activity during work hours and specific periods of exercise. The type of specific physical activity and weekly frequency were reported at each of the body composition visits. In addition, daily pedometer readings, measuring steps per day, were recorded for 2 weeks before the assessments. The specific physical activity intensity, session time, and session frequency product and any change in product were used as a continuous variable in assessing influence of body composition. Pedometer recordings were examined as both continuous and binary variables with groups divided according to above or below the median count or change in count. Attendance for body composition analysis was a voluntary aspect of the study. Although the patients were encouraged to attend, reasons for non-attendance included loss to follow-up, lack of time, distance, and pregnancy. Statistical Analysis Only those patients who completed all of the body composition studies were included within the analysis. Data are presented as mean ⫾ standard deviation. All body composition variables at each time interval provided an acceptable normal distribution (kurtosis and skewness ⬎ ⫺2 and ⬍2.0), and there was no significance difference in variance between groups (Levene’s test) for any variable. Withingroup changes were assessed using paired Student’s t tests, and between-group changes were assessed using unpaired

Student’s t tests. Univariate correlations were assessed using Pearson correlation coefficients. Analyses at baseline and 6 months and baseline and 24 months were examined separately. Linear regression analysis was used to assess the effect of multiple variables for influence on the variance of outcome measures. All analysis was performed using SPSS 14 (SPSS, Inc., Chicago, IL).

Results There were 80 patients recruited into the study, 40 into each treatment program. The complete set of body composition studies were performed on 32 surgical and 40 VLED/ medical subjects at baseline. Measurements were repeated at 6 and 24 months on 29 and 26 surgical subjects and 32 and 27 VLED/medical subjects, respectively. Thus, 85% of possible subjects attended at 6 months and 74% attended at 24 months. The weight loss for those who attended all body composition analyses was not different from those who did not attend. Measurement of total body weight by scales and DXA differed significantly, with DXA providing a slightly lower weight, particularly before weight loss. DXA underestimated weight by 1.4 ⫾ 1.0 kg [95% confidence interval (CI), 1.1 to 1.7 kg) at baseline, by 0.4 ⫾ 0.5 kg (95% CI, 0.27 to 0.54 kg; p ⬍ 0.001 for both) at 6 months, and by 0.32 ⫾ 0.8 kg (95% CI, 0.12 to 0.53 kg; p ⬍ 0.005 for all) at 24 months. There was a greater underestimation in the higher-weight subjects, with a significant negative correlation between DXA weight–scale weight difference and the scale weight (r ⫽ ⫺0.30, p ⫽ 0.03). There were no differences between the surgical and VLED/medical groups in this discrepancy at baseline and 6 months, but at 24 months, the surgical group had a smaller discrepancy related to their lower mean weight (p ⫽ 0.01). Thus, DXA weight accurately reflected weight within the 2% to 3% coefficient of variation expected for DXA soft tissue, but it slightly underestimated weight compared with scales, and this underestimation was greater in the baseline Class 1 (BMI 30 to 35 kg/m2) obese subjects. Comparison of Baseline and 6 Months The characteristics of the 61 subjects with paired body composition analysis at baseline and 6 months are shown in Table 1. The groups were well matched for all measured characteristics at baseline, and weight loss and change in basic anthropometric measures at 6 months after randomization were very similar (Table 1). Changes in weight, abdominal circumference, and waist-to-hip ratio were all significant and similar for both groups at 6 months. Change in Total Body Fat, Fat-free Mass, and Skeletal Muscle Mass at 6 Months Baseline and 6-month DXA measures of total fat mass and fat-free mass for the two randomized groups are shown OBESITY Vol. 15 No. 5 May 2007

1189

Body Composition with Weight Loss, Dixon et al.

Table 1. Baseline characteristics of subjects (n ⫽ 61) who had body composition studies before and 6 months after randomization to surgical and VLED/medical weight loss therapy Surgical (LAGB) Number Age (years) Percent men BMI (kg/m2) Weight (kg) WHO abdominal circumference (cm) Maximum gluteal circumference (cm) Waist-to-hip ratio* Weight at 6 months (kg) Weight loss at 6 months (kg) Change in abdominal circumference (cm) Change in waist-to-hip ratio*

29 41.8 ⫾ 6. 24.1% 33.6 ⫾ 1.6 95.8 ⫾ 11.3 103.1 ⫾ 8.2 118.4 ⫾ 6.3 0.874 ⫾ 0.09 81.7 ⫾ 11.5 14.1 ⫾ 4.5 ⫺11.1 ⫾ 4.7 ⫺0.023

VLED/medical 32 40.6 ⫾ 7.2 25% 33.3 ⫾ 1.3 93.3 ⫾ 9.9 101.7 ⫾ 7.4 116.5 ⫾ 6.3 0.876 ⫾ 0.09 80.0 ⫾ 11.0 13.3 ⫾ 7.2 ⫺11.5 ⫾ 7.3 ⫺0.032

p 0.47 0.94 0.34 0.35 0.47 0.23 0.93 0.55 0.60 0.80 0.49

WHO, World Health Organization. Values are mean ⫾ SD, unpaired Student’s t test. ␹2 test for percent men. Changes in weight, abdominal circumference, and waist-to-hip ratio were all significant for both groups at 6 months. * Waist-to-hip ratio, WHO abdominal circumference/maximal gluteal circumference.

in Table 2. Weight loss was associated with significant decreases in fat mass and fat-free mass in both groups (Table 2). Similar weight loss (Table 1) was associated with similar loss of fat mass, but the surgical group lost more fat-free mass and had a higher proportion of fat-free mass to fat mass loss at 6 months. Expected proportions of fat-fee mass loss to fat mass loss, based on the Gallagher formula, were 1:3.5 for both groups at 6 months (15). Weight loss in the surgical group followed the expected ratio of 1:3.5, but for the VLED/medical group, there was a lower loss of fat-free mass, with a fat-free mass loss to fat mass loss ratio of 1:8.5. At 6 months, there was no evidence of preferential trunk fat loss compared with limb fat loss in either group. The percentage change in fat content in the limbs and trunk was similar to the change in total body fat (Figure 1). There was no change in limb to trunk fat ratio in either group at 6 months (data not shown). Reported exercise and pedometer recordings did not significantly influence any change in the trunk to limb fat ratio at 6 months. There was a statistically significant loss in mean skeletal muscle mass as measured by DXA in the surgical group (Table 2). The surgical group had a higher proportion of skeletal muscle mass loss to total fat mass loss at 6 months, but the ratio was quite low (1:7.2). At 6 months, 18 patients reported a reduction or no change in physical activity since the study start, and 22 recorded a reduction or no change in pedometer scores. 1190

OBESITY Vol. 15 No. 5 May 2007

Neither the reported exercise activity at baseline nor the change in activity at 6 months was associated with improved weight loss or influenced fat-free mass loss in either group. Similarly, baseline pedometer scores and change in pedometer scores at 6 months did not influence weight loss or fat-free mass loss in either group. Change in Total Body Bone Mineral Content, Total Body Potassium, and Total Body Nitrogen at 6 Months There were no mean changes in total bone mineral content in either group at 6 months. With both forms of weight loss, there were similar small but significant decreases in total body potassium, potassium index, and body cell mass (Table 2). Both total body protein and nitrogen index increased significantly with weight loss in the VLED/medical group but did not change in the surgical group. Despite this small significant rise in the VLED/medical group, there were no statistically significant differences between the groups in either total body protein or nitrogen index at baseline and 6 months. The body protein for each kilogram of body cell mass was significantly higher in both groups at 6 months after randomization. For the VLED/medical group, measures were 269 ⫾ 26 and 296 ⫾ 30 g/kg (p ⬍ 0.001) and for the surgical group were 262 ⫾ 23 and 282 ⫾ 34 g/kg (p ⫽ 0.002) at baseline and 6 months, respectively. There were no differences between the two groups at baseline or at 6 months.

Body Composition with Weight Loss, Dixon et al.

Table 2. Body composition studies before (T0) and at 6 months (T6) after randomization to surgical and VLED/medical weight loss programs

DXA weight loss (kg) T0 fat-free mass (DXA) T6 fat-free mass (DXA) ⌬ fat-free mass (kg) p value T0 total fat mass (DXA) T6 total fat mass (DXA) ⌬ total fat mass (kg) p value ⌬ fat-free mass:total fat mass ratio T0 total body bone mineral content (kg) T6 total body bone mineral content (kg) ⌬ total body bone mineral content (kg) p value T0 skeletal muscle mass (kg) T6 skeletal muscle mass (kg) ⌬ skeletal muscle mass (kg) p value ⌬ skeletal muscle:total fat mass ratio T0 total body potassium (g) T6 total body potassium (g) ⌬ total body potassium (g) p value T0 potassium index T6 potassium index ⌬ potassium index p value T0 body cell mass (kg) T6 body cell mass (kg) ⌬ body cell mass p value T0 total body protein (kg) T6 total body protein (kg) ⌬ total body protein p value T0 nitrogen index T6 nitrogen index ⌬ nitrogen index p value

Surgical (n ⴝ 29)

VLED/medical (n ⴝ 32)

⫺12.9 ⫾ 4.0 50.9 ⫾ 11.3 48.1 ⫾ 11.0 ⫺2.9 ⫾ 2.0 ⬍0.001* 43.3 ⫾ 7.4 33.2 ⫾ 7.9 ⫺10.1 ⫾ 3.3 ⬍0.001* 1:3.5 3.10 ⫾ 0.41 3.08 ⫾ 0.42 ⫺0.015 ⫾ 0.11 0.50* 24.2 ⫾ 6.4 22.8 ⫾ 6.2 ⫺1.4 ⫾ 1.1 ⬍0.001* 1:7.2 170.8 ⫾ 35.8 161.5 ⫾ 36.8 ⫺11.6 ⫾ 14.5 ⬍0.001* 157.9 ⫾ 12.4 147.9 ⫾ 12.3 ⫺10.0 ⫾ 13.6 0.001* 36.4 ⫾ 7.7 34.5 ⫾ 7.8 ⫺2.46 ⫾ 3.1 ⬍0.001* 9.59 ⫾ 2.49 9.60 ⫾ 2.50 0.01 ⫾ 0.55 0.96* 1.03 ⫾ 0.1 1.04 ⫾ 0.1 0.007 ⫾ 0.06 0.54*

⫺12.4 ⫾ 7.3 50.4 ⫾ 9.6 49.2 ⫾ 9.4 ⫺1.3 ⫾ 1.7 ⬍0.001* 41.5 ⫾ 6.7 30.4 ⫾ 8.3 ⫺11.4 ⫾ 6.8 ⬍0.001* 1:8.5 3.08 ⫾ 0.43 3.07 ⫾ 0.43 ⫺0.012 ⫾ 0.09 0.50* 24.1 ⫾ 5.5 23.5 ⫾ 5.1 ⫺0.6 ⫾ 1.1 0.003* 1:19 173.4 ⫾ 31.2 160.9 ⫾ 31.6 ⫺10.1 ⫾ 13.7 ⬍0.001* 158.1 ⫾ 14.3 148.9 ⫾ 13.2 ⫺9.1 ⫾ 13.1 0.001* 37.1 ⫾ 6.7 34.4 ⫾ 6.7 ⫺2.17 ⫾ 2.9 ⬍0.001* 9.92 ⫾ 2.11 10.18 ⫾ 1.99 0.26 ⫾ 0.63 0.027* 1.06 ⫾ 0.1 1.10 ⫾ 0.1 0.033 ⫾ 0.07 0.008*

p 0.72 0.86 0.67 0.002 0.31 0.18 0.44

0.84 0.86 0.90 0.97 0.64 0.01 0.742 0.77 0.94 0.68 0.95 0.58 0.81 0.76 0.94 0.72 0.58 0.32 0.11 0.18 0.18 0.12

Values are mean ⫾ SD, unpaired Student;s t test for between-group differences. *Values are mean ⫾ SD, paired Student’s t test for intragroup changes.

OBESITY Vol. 15 No. 5 May 2007

1191

Body Composition with Weight Loss, Dixon et al.

Figure 1: The change in the percentage of fat lost from trunk compared with fat lost from limbs as measured by DXA between baseline and 6 months.

Comparison of Baseline and 2 Years The characteristics of the 53 patients who completed paired baseline and 2-year body composition analyses are presented in Table 3. At 2 years, the surgical group had lost considerably more weight and had more favorable waist and waist-to-hip ratio measures. DXA underestimated weight change by ⬃10% compared with scale weight change, and this underestimation was greater with greater weight loss (r ⫽ ⫺0.46, p ⬍ 0.001). This is related to the DXA underestimation of weight in obese subjects. At 2 years, the mean weight change for the two groups combined was ⫺13.0 ⫾ 10.1 kg, with a range from ⫹11.6 to ⫺32.0 kg providing a broad range of weight change.

Change in Total Body Fat, Fat-free Mass, and Skeletal Muscle Mass at 2 Years In association with greater weight loss, the surgical group had significantly greater decreases in both total body fat and fat-free mass, but the ratio of the change in fat-free mass to fat mass was very similar (surgical group, 1:5.5; VLED/ medical group, 1:5.9). These ratios are favorable compared with differences expected using the Gallagher formula for percentage of fat mass in weight-stable individuals at mean baseline and 2-year BMI. Expected proportions of fat-free mass loss to fat mass loss were 1:3.5 and 1:3.3 for the surgical and VLED/medical groups, respectively (15). There was no evidence of preferential trunk fat loss compared with limb fat loss in either group (Figure 2), and the ratio of trunk to limb fat remained remarkably constant between baseline and 2 years (r ⫽ 0.96, p ⬍ 0.001; Figure 3). Using linear regression analysis, the change in limb to trunk fat ratio was not influenced by the extent of weight change, sex, reported exercise, or treatment group. There was, however, an influence based on pedometer scores. Those who recorded a greater than the median increase in steps per day of 1500 steps (⫺0.05, p ⫽ 0.02) or at 24 months had greater than the median number of steps per day of 9500 had a decrease in the trunk to limb fat ratio (⫺0.053, p ⫽ 0.01). Although significant, the effect was small, with the trunk to limb fat ratio at baseline accounting for 92% (p ⬍ 0.001) of variance, and an increase of 1500 steps per day adding 2% (p ⫽ 0.008) of variance of trunk to limb fat ratio at 2 years. These two factors had an indepen-

Table 3. Baseline characteristics of subjects (n ⫽ 53) who had body composition studies before and 24 months after randomization into either surgical or VLED/medical weight loss therapy

Number Age (years) Percent men BMI (kg/m2) Weight (kg) WHO abdominal circumference (cm) Maximum gluteal circumference (cm) Waist-to-hip ratio* Weight at 24-months (kg) Weight loss at 24-months (kg) Change in abdominal circumference (cm) Change in waist-to-hip ratio*

Surgical (LAGB)

VLED/medical

p

26 42.2 ⫾ 6.4 26% 33.6 ⫾ 1.6 95.2 ⫾ 11.7 102.7 ⫾ 8.5 118.2 ⫾ 6.4 0.872 ⫾ 0.096 74.9 ⫾ 11.5 20.3 ⫾ 6.5 ⫺15.3 ⫾ 5.7 ⫺0.033

27 40.6 ⫾ 6.9 23% 33.2 ⫾ 1.3 93.3 ⫾ 9.7 101.8 ⫾ 7.0 116.2 ⫾ 6.2 0.879 ⫾ 0.088 87.4 ⫾ 11.2 5.9 ⫾ 8.0 ⫺3.8 ⫾ 7.2 ⫺0.0042

0.47 0.81 0.34 0.35 0.47 0.23 0.93 ⬍0.001 ⬍0.001 ⬍0.001 0.01

WHO, World Health Organization. Values are mean ⫾ SD, unpaired Student’s t test. ␹2 test for percent men. *Waist-to-hip ratio, WHO abdominal circumference/maximal gluteal circumference.

1192

OBESITY Vol. 15 No. 5 May 2007

Body Composition with Weight Loss, Dixon et al.

Figure 2: The change in the percentage of fat lost from trunk compared with fat lost from limbs as measured by DXA between baseline and 24 months.

dent effect, and walking ⬎9500 steps a day provided no additional effect. When changes in trunk fat and limb fat were examined separately, the effect of increasing the number of steps per day was limited to a small reduction in trunk fat rather than an increase in limb fat when controlled for the subject’s weight loss. There was a greater decrease in skeletal muscle mass in the surgical group, but a similar ratio of skeletal muscle mass loss to fat mass loss ratio in both groups (Table 4). Using linear regression to analyze the two groups together, we found that the change in weight accounted for the change in anthropometric measures, fat mass, fat-free mass, limb fat, trunk fat, and skeletal muscle mass. There was no additional variance attributable to the treatment group. Similar to the analysis at 6 months, there was no significant relationship between weight loss and fat-free mass loss and reported exercise or pedometer records at 24 months. Change in Total Body Bone Mineral Content, Total Body Potassium, and Total Body Nitrogen at 2 Years In association with greater weight loss at 2 years, the surgical group had lower total body potassium, lower potassium index, and a reduction in body cell mass. With the two groups combined, there was a strong positive correlation between weight change and change in body cell mass (r ⫽ 0.72, p ⬍ 0.001), and after controlling for weight change, the group to which a patient was randomized did not influence the change in body cell mass. There were no differences between the groups in total body bone mineral content at baseline and 2 years. There was a small but significant decrease in total bone mineral content in both groups at 2 years (2.8% in the surgical group and 2% in the VLED/medical group), but with the groups combined, there was no significant correlation between change in bone mineral content and change in weight, body cell mass, total body protein, fat-free mass, or randomized treatment group. There was no correlation between change in bone mineral content and age or sex, but there was an age

Figure 3: The ratio of trunk fat to limb fat at baseline and at 2 years was remarkably similar for each individual (r ⫽ 0.96, p ⬍ 0.001).

cut-off of 50 years for entry into the study. Reported exercise and recorded steps per day did not influence the change in total bone mineral content. Taken with the baseline and 6-month data, it seems unlikely that the small decrease in bone mineral content was significantly influenced by weight loss. There was no significant change in nitrogen index in either group, but a small decrease in total body protein was found in the surgical group. There was a 3.3 ⫾ 6.9% loss in total body protein, with a 21.4 ⫾ 6.4% loss in body weight, for a ratio of 1:6.5 for the surgical group. Figures for the VLED/medical group were 1.4 ⫾ 8.4%, 6.2 ⫾ 8.5%, and 1:4.3, respectively. Regression analysis performed with the two groups combined showed no significant change in total body protein or nitrogen index with the change in weight or fat-free mass over the 2-year period, and the group to which the patient was randomized had no statistical effect. The body protein for each kilogram of body cell mass was significantly higher in the surgical compared with the VLED/medical group at 24 months (p ⫽ 0.007). In paired analysis of total body protein for each kilogram of body cell mass, there was no change in the VLED/medical group (268 ⫾ 25 g/kg at baseline; 269 ⫾ 27 g/kg at 24 months, p ⫽ 0.86), but a significant increase in the surgical group (261 ⫾ 24 g/kg at baseline; 292 ⫾ 34 g/kg at 24 months, p ⬍ 0.001) was seen. With the two groups combined, we found a strong negative relationship between weight loss and the protein per kilogram of body cell mass. The grouping at randomization had no influence (Figure 4). There was no change in protein per kilogram of fat-free mass in either OBESITY Vol. 15 No. 5 May 2007

1193

Body Composition with Weight Loss, Dixon et al.

Table 4. Body composition studies before (T0) and at 24 months (T24) after randomization into surgical or VLED/medical weight loss programs

DXA weight loss T0 fat-free mass (DXA) T24 fat-free mass (DXA) ⌬ fat-free mass (kg) p value T0 total fat mass (DXA) T24 total fat mass (DXA) ⌬ total fat mass (kg) p value ⌬ fat-free mass:total fat mass ratio T0 total body bone mineral content (kg) T24 total body bone mineral content (kg) ⌬ total body bone mineral content (kg) p value T0 skeletal muscle mass (kg) T24 skeletal muscle mass (kg) ⌬ skeletal muscle mass (kg) p value ⌬ skeletal muscle:total fat mass ratio T0 total body potassium (g) T24 total body potassium (g) ⌬ total body potassium (g) p value T0 potassium index T24 potassium Index ⌬ potassium index p value T0 body cell mass (kg) T24 body cell mass (kg) ⌬ body cell mass (kg) p value T0 total body protein (kg) T24 total body protein (kg) ⌬ total body protein (kg) p value T0 nitrogen index T24 nitrogen index ⌬ nitrogen index p value

Surgical (n ⴝ 26)

VLED/medical (n ⴝ 27)

⫺18.8 ⫾ 6.1 50.4 ⫾ 11.5 47.5 ⫾ 11.0 ⫺2.9 ⫾ 2.2 ⬍0.001* 43.3 ⫾ 7.7 27.4 ⫾ 7.7 ⫺15.9 ⫾ 5.7 ⬍0.001* 1:5.5 3.08 ⫾ 0.43 3.00 ⫾ 0.44 0.087 ⫾ 0.12 0.002* 23.9 ⫾ 6.5 22.6 ⫾ 6.1 ⫺1.3 ⫾ 1.1 ⬍0.001* 1:12.2 169.3 ⫾ 36.3 147.0 ⫾ 34.6 ⫺24.5 ⫾ 16.7 ⬍0.001* 158.3 ⫾ 12.3 137.3 ⫾ 12.4 ⫺20.6 ⫾ 14.8 ⬍0.001* 36.1 ⫾ 7.7 31.7 ⫾ 7.4 ⫺4.4 ⫾ 3.3 ⬍0.001* 9.46 ⫾ 2.51 9.11 ⫾ 2.32 ⫺0.35 ⫾ 0.71 0.02* 1.03 ⫾ 0.11 1.01 ⫾ 0.10 0.02 ⫾ 0.0 0.14*

⫺5.1 ⫾ 7.7 50.6 ⫾ 10.1 49.8 ⫾ 9.6 ⫺0.75 ⫾ 1.9 0.053* 41.4 ⫾ 6.7 37.1 ⫾ 8.6 ⫺4.4 ⫾ 6.9 0.003* 1:5.9 3.07 ⫾ 0.41 3.01 ⫾ 0.43 0.061 ⫾ 0.9 0.002* 24.2 ⫾ 5.7 23.7 ⫾ 5.3 ⫺0.4 ⫾ 1.6 0.06* 1:11.0 173.6 ⫾ 30.9 170.3 ⫾ 31.5 ⫺2.5 ⫾ 10.6 0.24* 159.4 ⫾ 14.7 158.5 ⫾ 14.5 ⫺1.6 ⫾ 10.7 0.47* 37.1 ⫾ 6.6 36.4 ⫾ 6.7 ⫺0.7 ⫾ 2.2 0.23* 9.93 ⫾ 2.17 9.73 ⫾ 1.99 ⫺0.19. ⫾ 0.85 0.28* 1.06 ⫾ 0.08 1.05 ⫾ 0.09 0.01 ⫾ 0.0 0.55*

Values are mean ⫾ SD, unpaired Student’s t test for between-group differences. *Values are mean ⫾ SD, paired Student’s t test for intragroup changes.

1194

OBESITY Vol. 15 No. 5 May 2007

p ⬍0.001 0.95 0.42 ⬍0.001 0.34 ⬍0.001 ⬍0.001

0.93 0.89 0.39 0.87 0.48 0.009

0.61 0.013 ⬍0.001 0.78 ⬍0.001 ⬍0.001 0.64 0.02 ⬍0.001 0.47 0.45 0.49 0.25 0.11 0.62

Body Composition with Weight Loss, Dixon et al.

Figure 4: With weight loss, there was a significant increase in protein (grams) per kilogram of body cell mass as determined by total body nitrogen (r ⫽ ⫺0.56, p ⬍ 0.001). Group mean ⫾ 95% CI is also shown. The surgical group had a significantly greater increase (p ⫽ 0.001) in association with greater weight loss.

group with weight loss (data not shown). Reported exercise and pedometer recordings did not significantly influence any of the protein measures.

Discussion There were small, but statistically significant, differences between the two groups of patients at 6 months, when the mean weight loss achieved by each group was similar. The surgical group lost more fat-free mass and skeletal muscle mass as estimated by DXA. There are several possible explanations for the differences seen. A VLED that provides a moderate level of protein intake may protect against loss of fat-free mass. The patients having LAGB surgery may have had a greater reduction in dietary protein intake. The short period of postsurgical convalescence may also have had an effect. Differences in physical activity are an unlikely explanation, because both groups were given similar advice and support, reports of physical activity were not different, and there was no observable relationship between reported physical activity and change in fat-free mass. Importantly, for neither group was there a decrease in total body protein or nitrogen index, but an increase was found in

the medical group, suggesting a benefit associated with the moderate protein intake of the VLED. Although at 6 months there was no clinical concern regarding fat-free mass loss in the surgical group, any loss may be reduced by ensuring that dietary protein intake is adequate. Loss of fat mass at both 6 months and 2 years was similar from both trunk and limb stores. There was no evidence of preferential trunk loss in association with weight loss. There was a remarkably constant trunk to limb fat ratio for each individual, and the ratio was not influenced by the extent of weight loss, sex, or method of weight loss. There was a small but significant reduction in the ratio of central to limb fat in those who recorded an increase of ⱖ1500 steps/d. This study did not look specifically at intra-abdominal or visceral fat changes and did not measure changes immediately after rapid weight loss, when visceral loss may have been preferentially greater (16 –20). However, although insulin sensitivity is closely related to fat distribution in an individual, improvement in insulin sensitivity with weight loss is more closely related to total abdominal fat loss than specifically to visceral fat (17). Although preferential visceral fat loss is often reported (21), the rigor of methodology varies, and preferential visceral fat loss with weight loss over a short period may simply reflect a transient initial effect that corrects over time. Most studies that we found describing a preferential effect were short-term studies generally of up to 6 months’ duration. Our findings at 2 years suggest that the ratio of trunk to limb fat is quite constant and that the aim of weight loss in the long term should be total fat loss. In addition, increased physical activity may provide a subtle, but significant, increased loss of central fat. At 2 years, the ratio of fat-free mass loss to fat mass loss was similar in both groups (1:5.7) and more favorable than that expected (1:3.4) for the change in BMI achieved. We have previously reported these favorable changes in a smaller observational study of LAGB subjects (22), and our findings are similar to other reports after LAGB surgery (23,24). These contrast with a greater proportion of fat-free mass loss after malabsorptive bariatric surgical procedures (25,26) and emphasize the nutritional and metabolic safety of non-malabsorptive procedures. Roux-en-Y gastric bypass is currently the most commonly used procedure in the world, but data regarding fat-free mass loss are scant and unfavorable. In the medium term, the weight loss after gastric bypass and LAGB is comparable (27). It is, therefore, of concern that, for the same weight loss, there is likely to be greater fat-free mass loss and lower fat mass loss after gastric bypass. We have expressed DXA assessment of bone only as bone mineral content, because there may be a significant artifact effect when assessing bone area and calculating areal bone density with weight loss (28,29). We found such an inconsistency in this study, because total bone mineral density did not change, but age-matched bone OBESITY Vol. 15 No. 5 May 2007

1195

Body Composition with Weight Loss, Dixon et al.

mineral density rose significantly, whereas total bone mineral content fell slightly between baseline and 2 years in both groups (data not shown). In our study, we suspect that bone area is overestimated in obese subjects. It is very reassuring that, in this study, there was no major decrease in total bone mineral content, and any changes that occurred do not seem to have been influenced by the method used to achieve weight loss or by the extent of weight loss. It is possible that the small decrease in total bone mineral content is largely related to aging over 2 years, because the majority of patients were women with a mean age of 41 years (30,31). It should also be recognized that there is a positive relationship between BMI and age-adjusted bone mineral content (32), and there may be a trend to normative reduction in total bone mineral content with a significant fall in BMI. However, we would have expected a differential effect with weight loss. Bone mineral content is more closely related to fat-free mass than to fat mass (33). A longer follow-up of these subjects may assist in separating the influences of aging and weight loss. The small decrease in bone mineral content contrasts with changes in bone content and metabolism after Roux-en Y gastric bypass (34 –36) and biliopancreatic diversion (37,38), which are procedures that predictably alter calcium absorption and may also influence vitamin D absorption. At 2 years after randomization, there was no relationship between the change in total body protein and weight loss with groups combined. This would seem quite remarkable considering the significant change in weight and the strong relationship between weight loss and change in both fat-free mass and body cell mass. In fact, there was a weight loss–related increase in protein per kilogram of body cell mass and no change in the protein per kilogram of fat-free mass. These findings suggest that the intracellular density of protein increases with weight loss and may reflect structural and functional changes in muscle with weight loss (39). This shows remarkable safety with regard to protein nutrition for both treatment groups. The very low loss of fat-free mass and protein conservation in the VLED/medical program contrast with reports of significant fat-free mass loss after other programs using VLEDs (40). There are several limitations to this study. We did not measure total body water and cannot make assumption regarding body hydration. Neither treatment group was given the specified treatment program alone. All subjects received regular support and advice regarding eating behaviors, diet, multivitamin supplementation, and physical activity throughout the 2-year period of the study. Thus, our findings may not necessarily be applicable to programs with less intense follow-up. The study followed patients with Class I (BMI, 30 to 35 kg/m2) obesity, and results may not necessarily apply to those with Classes II 1196

OBESITY Vol. 15 No. 5 May 2007

(BMI, 35 to 40 kg/m2) and III (BMI, ⬎40 kg/m2) who generally lose greater amounts of weight after bariatric surgery and VLED therapy. Although we followed a large percentage of eligible subjects, there remains the possibility that those not returning for repeat studies were less compliant and may have had less favorable body composition outcomes. We did not measure energy expenditure but obtained patient-generated questionnaire and pedometer data regarding physical activity. Neither baseline activity nor the change in reported physical activity influenced weight loss or any of the measures of fat-free body composition in either group throughout the study period. There was the subtle effect at 2 years in fat distribution in those increasing the number of steps per day. The data regarding exercise protecting against loss of fat-free mass with both VLED and bariatric surgery are limited and not convincing (41,42). Quality studies looking at the duration and nature of physical activity in association with substantial weight loss are needed. It is encouraging that the extensive and sustained weight loss in the surgical group after LAGB surgery seems to be favorable with regard to all parameters measured in these mild to moderately obese subjects. The adjustability of the LAGB provides an ability to achieve satiety and avoid excessive obstruction (43), allowing a reduction in quantity, but not quality, of food intake. In addition, there is no gastrointestinal diversion that puts at risk micronutrient and macronutrient absorption. We also showed that combining VLED and Orlistat as a 6-month weight loss program provides excellent weight loss and body composition outcomes, but the maintenance of the weight loss over the medium term is problematic, as shown in many previous studies. This weight regain may have been mitigated if sibutramine had been used after the VLED therapy, but this medication was not available in Australia at the time of the study.

Acknowledgments The authors thank Margaret Anderson for data entry and data management of this study; Dr. C. Sam Lo and Dr. Daniel B. Stroud of the Monash Medical Center, Body Composition Laboratory, for assistance and expertise in measuring total body nitrogen and potassium; and Dr. Veronica Alverez for intellectual assistance in analysis of data and help with manuscript preparation. This study was supported by Inamed Health, manufacturer of the LAP-BAND System; Novartis, manufacturer of Optifast; and U.S. Surgical Corporation, manufacturer of disposable laparoscopic instruments. References 1. Dixon JB, O’Brien PE. Changes in comorbidities and improvements in quality of life after LAP-BAND placement. Am J Surg. 2002;184:S51– 4.

Body Composition with Weight Loss, Dixon et al.

2. Li Z, Maglione M, Tu W, et al. Meta-analysis: pharmacologic treatment of obesity. Ann Intern Med. 2005;142:532– 46. 3. Anderson JW, Konz EC, Frederich RC, Wood CL. Longterm weight-loss maintenance: a meta-analysis of US studies. Am J Clin Nutr. 2001;74:579 – 84. 4. Chapman A, Kiroff G, Game P, et al. Laparoscopic adjustable gastric banding in the treatment of obesity: a systematic review. Surgery. 2004;135:326 –51. 5. Maggard MA, Shugarman LR, Suttorp M, et al. Metaanalysis: surgical treatment of obesity. Ann Intern Med. 2005; 142:547–59. 6. Dao HH, Frelut ML, Oberlin F, Peres G, Bourgeois P, Navarro J. Effects of a multidisciplinary weight loss intervention on body composition in obese adolescents. Int J Obes Relat Metab Disord. 2004;28:290 –9. 7. Das SK, Roberts SB, Kehayias JJ, et al. Body composition assessment in extreme obesity and after massive weight loss induced by gastric bypass surgery. Am J Physiol Endocrinol Metab. 2003;284:E1080 – 8. 8. O’Brien PE, Dixon JB, Laurie C, et al. Treatment of mild to moderate obesity with laparoscopic adjustable gastric banding or an intensive medical program: a randomized trial. Ann Intern Med. 2006;144:625–33. 9. O’Brien PE, Dixon JB, Brown W, et al. The laparoscopic adjustable gastric band (Lap-Band): a prospective study of medium-term effects on weight, health and quality of life. Obes Surg. 2002;12:652– 60. 10. Favretti F, O’Brien PE, Dixon JB. Patient management after LAP-BAND placement. Am J Surg. 2002;184:S38 – 41. 11. Kim J, Wang Z, Heymsfield SB, Baumgartner RN, Gallagher D. Total-body skeletal muscle mass: estimation by a new dual-energy X-ray absorptiometry method. Am J Clin Nutr. 2002;76:378 – 83. 12. Bruce A, Andersson M, Arvidsson B, Isaksson B. Body composition. Prediction of normal body potassium, body water and body fat in adults on the basis of body height, body weight and age. Scand J Clin Lab Invest. 1980;40:461–73. 13. Wang Z, St-Onge MP, Lecumberri B, et al. Body cell mass: model development and validation at the cellular level of body composition. Am J Physiol Endocrinol Metab. 2004;286: E123– 8. 14. Stroud DB, Borovnicar DJ, Lambert JR, et al. Clinical studies of total body nitrogen in an Australian hospital. Basic Life Sci. 1990;55:177– 82. 15. Gallagher D, Visser M, Sepulveda D, Pierson RN, Harris T, Heymsfield SB. How useful is body mass index for comparison of body fatness across age, sex, and ethnic groups? Am J Epidemiol. 1996;143:228 –39. 16. Tiikkainen M, Bergholm R, Rissanen A, et al. Effects of equal weight loss with orlistat and placebo on body fat and serum fatty acid composition and insulin resistance in obese women. Am J Clin Nutr. 2004;79:22–30. 17. Laaksonen DE, Kainulainen S, Rissanen A, Niskanen L. Relationships between changes in abdominal fat distribution and insulin sensitivity during a very low calorie diet in abdominally obese men and women. Nutr Metab Cardiovasc Dis. 2003;13:349 –56.

18. Marks SJ, Chin S, Strauss BJ. The metabolic effects of preferential reduction of visceral adipose tissue in abdominally obese men. Int J Obes Relat Metab Disord. 1998;22: 893– 8. 19. Busetto L, Tregnaghi A, Di Marchi F, et al. Liver volume and visceral obesity in women with hepatic steatosis undergoing gastric banding. Obes Res. 2002;10:408 –11. 20. Thomas EL, Brynes AE, McCarthy J, et al. Preferential loss of visceral fat following aerobic exercise, measured by magnetic resonance imaging. Lipids. 2000;35:769 – 76. 21. Smith SR, Zachwieja JJ. Visceral adipose tissue: a critical review of intervention strategies. Int J Obes Relat Metab Disord. 1999;23:329 –35. 22. Strauss BJ, Marks SJ, Growcott JP, et al. Body composition changes following laparoscopic gastric banding for morbid obesity. Acta Diabetol. 2003;40 Suppl 1:S266 –9. 23. Pugnale N, Giusti V, Suter M, et al. Bone metabolism and risk of secondary hyperparathyroidism 12 months after gastric banding in obese pre-menopausal women. Int J Obes Relat Metab Disord. 2003;27:110 – 6. 24. Giusti V, Suter M, Heraief E, Gaillard RC, Burckhardt P. Effects of laparoscopic gastric banding on body composition, metabolic profile and nutritional status of obese women: 12months follow-up. Obes Surg. 2004;14:239 – 45. 25. Muscelli E, Mingrone G, Camastra S, et al. Differential effect of weight loss on insulin resistance in surgically treated obese patients. Am J Med. 2005;118:51–7. 26. Mingrone G, Greco AV, Giancaterini A, Scarfone A, Castagneto M, Pugeat M. Sex hormone-binding globulin levels and cardiovascular risk factors in morbidly obese subjects before and after weight reduction induced by diet or malabsorptive surgery. Atherosclerosis. 2002;161:455– 62. 27. O’Brien PE, McPhail T, Chaston TB, Dixon JB. Systematic review of medium-term weight loss after bariatric operations. Obes Surg. 2006;16:1032– 40. 28. Van Loan MD, Johnson HL, Barbieri TF. Effect of weight loss on bone mineral content and bone mineral density in obese women. Am J Clin Nutr. 1998;67:734 – 8. 29. Fogelholm GM, Sievanen HT, Kukkonen-Harjula TK, Pasanen ME. Bone mineral density during reduction, maintenance and regain of body weight in premenopausal, obese women. Osteoporos Int. 2001;12:199 –206. 30. Lakatos P, Krasznai I, Vargha P, Gergely I, Hollo I. Changes with aging in the bone mineral content of the lumbar spine and femoral neck in healthy women in Hungary. Acta Med Hung. 1989;46:73– 83. 31. Shipman AJ, Guy GW, Smith I, Ostlere S, Greer W, Smith R. Vertebral bone mineral density, content and area in 8789 normal women aged 33–73 years who have never had hormone replacement therapy. Osteoporos Int. 1999;9: 420 – 6. 32. Revilla M, Villa LF, Sanchez-Atrio A, Hernandez ER, Rico H. Influence of body mass index on the age-related slope of total and regional bone mineral content. Calcif Tissue Int. 1997;61:134 – 8. OBESITY Vol. 15 No. 5 May 2007

1197

Body Composition with Weight Loss, Dixon et al.

33. Aloia JF, Vaswani A, Ma R, Flaster E. To what extent is bone mass determined by fat-free or fat mass? Am J Clin Nutr. 1995;61:1110 – 4. 34. von Mach MA, Stoeckli R, Bilz S, Kraenzlin M, Langer I, Keller U. Changes in bone mineral content after surgical treatment of morbid obesity. Metabolism. 2004;53:918 –21. 35. Coates PS, Fernstrom JD, Fernstrom MH, Schauer PR, Greenspan SL. Gastric bypass surgery for morbid obesity leads to an increase in bone turnover and a decrease in bone mass. J Clin Endocrinol Metab. 2004;89:1061–5. 36. De Prisco C, Levine SN. Metabolic bone disease after gastric bypass surgery for obesity. Am J Med Sci. 2005;329:57– 61. 37. Marceau P, Biron S, Lebel S, et al. Does bone change after biliopancreatic diversion? J Gastrointest Surg. 2002;6:690 – 8. 38. Newbury L, Dolan K, Hatzifotis M, Low N, Fielding G. Calcium and vitamin D depletion and elevated parathyroid hormone following biliopancreatic diversion. Obes Surg. 2003;13:893–5.

1198

OBESITY Vol. 15 No. 5 May 2007

39. Kern PA, Simsolo RB, Fournier M. Effect of weight loss on muscle fiber type, fiber size, capillarity, and succinate dehydrogenase activity in humans. J Clin Endocrinol Metab. 1999; 84:4185–90. 40. Eston RG, Shephard S, Kreitzman S, et al. Effect of very low calorie diet on body composition and exercise response in sedentary women. Eur J Appl Physiol Occup Physiol. 1992; 65:452– 8. 41. Pronk NP, Donnelly JE, Pronk SJ. Strength changes induced by extreme dieting and exercise in extremely obese females. J Am Coll Nutr. 1992;11:152– 8. 42. Westerterp KR, Saris WH, Soeters PB, ten Hoor F. Determinants of weight loss after vertical banded gastroplasty. Int J Obes Relat Metab Disord. 1991;15:529 –34. 43. Dixon AF, Dixon JB, O’Brien PE. Laparoscopic adjustable gastric banding induces prolonged satiety: a randomized blind crossover study. J Clin Endocrinol Metab. 2005;90:813–9. Epub 2004 Dec 7.